arXiv:2607.10878v1 Announce Type: new
Abstract: AI agents are evolving from answer engines into persistent teams that use tools, delegate work, learn from experience, and modify the artifacts that shape their future behavior. The defining question for deployment is no longer merely what agents can do, but who controls what they are allowed to become. We introduce logos, a pluggable layer for self-evolution and governance that strengthens existing multiagent frameworks rather than replacing them. logos compiles heterogeneous multimodal inputs, including documents, images, audio, tables, databases, APIs, and human instructions into versioned agent packs containing agents, tools, knowledge, tests, permissions, and policies. During operation, it transforms agent activity into portable, auditable event traces and applies fail-closed verification across frameworks and backends. Every learned prompt, memory, skill, tool, role, or workflow remains an untrusted release candidate until held-out execution evidence, human-controlled policy, and explicit authorization permit its promotion. This architecture enables "verifiable human-agent loop engineering": agents can act, ask, learn, and propose improvements, while humans can steer objectives, permissions, approvals, and irreversible actions without interrupting continuous operation. logos provides a living logic for accountable automation. Agents may evolve at machine speed, but only evidence and human authority can close the loop.
Science Journals
arXiv:2607.10697v1 Announce Type: new
Abstract: We revisit Scheder's analysis of the original PPSZ algorithm. Keeping his regular and irregular estimates unchanged, we express them in common structural coordinates and replace only their final recombination by an explicit linear-programming dual certificate. The old and new running-time bounds are \[ \begin{array}{c|cc}
& \text{Unique-$3$-SAT} & \text{general $3$-SAT} \\ \hline \text{Scheder's analysis} & O^*(1.306972377^n) & O^*(1.307031594^n) \\ \text{this work} & O^*(1.306969598^n) & O^*(1.307031578^n). \end{array} \] In both rows, the general-case bound is obtained by applying the same existing Scheder--Steinberger unique-to-general lifting theorem to the corresponding Unique-$3$-SAT analysis. To the best of our knowledge, $O^*(1.307031578^n)$ is the best currently known worst-case randomized running-time bound for general $3$-SAT. Neither PPSZ nor the lifting theorem is modified. The numerical inequalities are certified by exact rational interval computation.
arXiv:2607.10136v1 Announce Type: new
Abstract: We examine anisotropy and spatial intermittency at small scales in drift-wave turbulence with zonal flows. We use a two-dimensional directional continuous wavelet transform, which allows simultaneous localization in scale, position, and direction. This wavelet analysis is applied to vorticity fields obtained from numerical simulations of the modified Hasegawa--Wakatani model, a reduced model of resistive drift-wave turbulence in magnetized plasmas with zonal flows. Directional wavelet statistics characterize the anisotropy of the turbulence. The second-order moment is enhanced around directions perpendicular to the zonal flow. Spatial intermittency, characterized by scale-dependent flatness, is more pronounced around directions along the zonal flow.
arXiv:2607.09694v1 Announce Type: new
Abstract: Attention Residuals replace the fixed residual sum with depthwise attention over previous sub-layer outputs in large language models (LLMs), but use each output as both a full-dimensional key and value. This couples routing with representation and makes depth-routing scores scale with the hidden width $d$. We propose Low-Rank Attention Residuals (LR-AttnRes), which keep full-dimensional residual values while using $r$-dimensional keys, with $r \ll d$, for routing. Projected LR-AttnRes emits learned low-rank keys from existing output projections, decoupling routing from residual content and achieving the best validation loss among the variants tested. Sliced LR-AttnRes uses the last $r$ dimensions of each value as the routing key, removing the auxiliary key-projection path and reducing residual-side FLOPs while still improving performance. Comprehensive sweeps show that depthwise routing can be effective with far fewer dimensions than the model width. We release code and models to facilitate future research.
arXiv:2607.10169v1 Announce Type: new
Abstract: Reinforcement learning (RL) has become a dominant paradigm for enhancing LLMs' reasoning capabilities. However, RL algorithms with PPO-Clip are inherently limited by exploration collapse. Subsequent works remain primarily heuristic and fail to identify the essential cause of PPO-Clip's failure. This work reveals the fundamental flaw of PPO-Clip: it implicitly measures policy discrepancy using Euclidean metric, which is theoretically inconsistent with the intrinsic geometry on the policy Riemannian manifold. This geometric mismatch results in overly conservative updates in low-probability regions while aggressive in high-probability regions, ultimately collapsing exploration. To correct this geometric flaw, we propose Riemannian Isometric Policy Optimization (RIPO), which guarantees isometric policy updates on the Riemannian manifold, effectively balancing exploration and exploitation. We further show that RIPO achieves a favorable bias-variance trade-off, which stabilizes optimization. Extensive experiments demonstrate that RIPO significantly surpasses existing LLM RL algorithms across seven competition-level benchmarks (up to 60% improvement over GRPO on AIME24).
arXiv:2607.10176v1 Announce Type: new
Abstract: Recently, the wave-particle turbulence simulation (WPTS) has been proposed as a novel framework for non-equilibrium turbulence modeling and simulation. In this work, for the first time the complete model equations of WPTS are explicitly derived from the perspective of wave-particle decomposition, and the physical mechanism of each term is clearly interpreted. To extend its applicability to wall-bounded flows, the WPTS coupled with wall model is developed, and the introduction of wall model substantially alleviates the near-wall grid-resolution constraint. In the bulk region, the wave component resolves the large-scale structures, whereas the particle component accounts for subgrid-scale modeling through the non-equilibrium transport mechanism. As a result, the coupled method enables accurate predictions of the flat-plate transition on coarse-grid. In particular, the computed skin-friction coefficient and mean velocity profiles in the fully turbulent region agree well with the reference data from direct numerical simulation, and the accuracy is markedly superior to that of the gas-kinetic scheme (GKS) under the identical grid. These findings underscore the considerable promise of the multi-scale WPTS method for transitional flow simulations.
arXiv:2607.10724v1 Announce Type: new
Abstract: We propose a four-field auxiliary reformulation of a time-discrete Cahn-Hilliard step. The construction is motivated by the scalar trace structure underlying two-dimensional Rafetseder-Zulehner decompositions and represents the scalar quantity $-\Delta c$ in the form $2p+div\,u$ through an auxiliary scalar field $p$ and an auxiliary vector field $u$. The resulting auxiliary problem is a mixed second-order Stokes/elasticity-type system, while the evolution equation for the phase field retains its standard mass-conserving gradient-flow structure. We derive a continuous four-field formulation that is equivalent to the classical convex-splitting mixed time step. We also state a conforming finite element discretization and prove one-step spatial estimates for the phase-field variables together with a stable auxiliary block estimate containing an explicit weak-Laplacian recovery defect. The numerical experiments verify the expected phase-field convergence in a manufactured setting, mass conservation, energy decay, and projected consistency of the auxiliary block.
arXiv:2505.23753v2 Announce Type: replace
Abstract: We introduce an approach for efficient Markov chain Monte Carlo (MCMC) sampling for challenging high-dimensional distributions in sparse Bayesian learning (SBL). The core innovation involves using hierarchical prior-normalizing transport maps (TMs), which are deterministic couplings that transform the sparsity-promoting SBL prior into a standard normal one. We analytically derive these prior-normalizing TMs by leveraging the product-like form of SBL priors and Knothe--Rosenblatt (KR) rearrangements. These transform the complex target posterior into a simpler reference distribution equipped with a standard normal prior that can be sampled more efficiently. Specifically, one can leverage the standard normal prior by using more efficient, structure-exploiting samplers. Our numerical experiments on various inverse problems -- including signal deblurring, inverting the non-linear inviscid Burgers equation, and recovering an impulse image -- demonstrate significant performance improvements for standard MCMC techniques.
arXiv:2607.11159v1 Announce Type: new
Abstract: Retrieval-Augmented Generation (RAG) significantly enhances the ability of Large Language Models (LLMs) to provide accurate and contextually relevant answers by dynamically integrating external databases. However, traditional RAG methods are primarily constrained by their reliance on text-based retrieval strategies, which often struggle with complex questions requiring multi-hop reasoning. To address this limitation, we introduce Neural Graph Matching based Retrieval-Augmented Generation (NGM-RAG), a novel framework that leverages graph structures to effectively capture and utilize relational knowledge for improved retrieval and answer generation. NGM-RAG explicitly incorporates graph construction, graph matching, and answer generation into a unified process. Within this framework, we propose a neural graph matching approach that combines text-based matching with Graph Neural Networks (GNNs). By employing an adaptive weighting strategy, NGM-RAG efficiently integrates multiple matching methods to select the most relevant contextual node information for answer generation. Experimental results on multi-hop question answering and long-context summarization tasks demonstrate that our NGM-RAG model achieves superior performance compared to both traditional NaiveRAG methods and state-of-the-art graph-enhanced approaches such as GraphRAG and LightRAG.
arXiv:2607.11164v1 Announce Type: new
Abstract: Joint communication and sensing has been a key technology in 6G. By integrating sensing into maritime communications, ships can communicate with the base station while sensing the surrounding environment to ensure safe navigation. In this paper, we introduce an integrated satellite-terrestrial maritime system (ISTMS) with joint communication and sensing based on the same radio-frequency signals. Specifically, the terrestrial base station (TBS) and low Earth orbit (LEO) satellite provide communication services for near-shore users (NSUs) and off-shore users (OSUs), respectively, while simultaneously performing target sensing. Based on a differential evolution method (DE), we propose a sensing algorithm, which can enhance the location accuracy and reduce resource consumption. Furthermore, we derive the key performance metrics for both communication and sensing. Through joint beamforming optimization of the TBS and LEO satellite, we maximize the sum rate of maritime users while satisfying target localization accuracy requirements and transmit power constraints. Finally, extensive simulation results demonstrate the effectiveness of the proposed algorithms in terms of location accuracy and transmission rate compared with the baseline algorithms.
arXiv:2607.10186v1 Announce Type: new
Abstract: Large language model (LLM) inference is increasingly limited by the capacity of High-Bandwidth Memory (HBM) in GPUs, as model weights and KV cache grow rapidly. High-Bandwidth Flash (HBF) provides higher capacity than HBM while retaining comparable bandwidth, making it a promising substrate for capacity-constrained LLM inference. However, its inherently high access latency, low bandwidth utilization, and lack of support for heterogeneous resource management make it difficult to integrate HBF into GPUs for LLM inference. We present FlashAccel, a co-designed system that enables efficient LLM inference using HBF. FlashAccel integrates HBF into HBM-based GPUs, providing architectural support to mitigate access latency. It improves bandwidth utilization through specialized data layouts for both model weights and KV cache, and introduces an HBF-aware storage management layer together with a programming model to organize persistent data in HBF and coordinate heterogeneous memory resources at the system level. Experimental results demonstrate that integrating six HBF stacks into the GPU enables FlashAccel to deliver an average improvement of 2.54$\times$ and 1.93$\times$ in throughput per GPU and energy efficiency over the HBM-only GPU under 100ms latency constraint, respectively.
arXiv:2607.10331v1 Announce Type: new
Abstract: Human-centered AI (HCAI) refers to guidelines or principles that aim on ethi-cally oriented design of systems. We compare HCAI- guidelines with princi-ples of socio-technical systems that emerged in the context of conventional in-formation technology. The comparison leads to a revision of socio-technical heuristics by including aspects of AI-usage. The comparison reveals that con-tinuous evolution is a basic characteristic of socio-technical systems, and that human oversight or interventions and the subsequent appropriation of AI-systems lead to continuous adaptation and re-design of the systems, if autono-my is collaboratively exercised. From a socio-technical point of view, the cru-cial requirement of transparency has not only to be fulfilled with technical fea-tures, but also by contributions of the whole system including human actors. It will be promising for using AI, if not only technical features, but organization-al and social practices are socio-technically designed in a way that compen-sates shortcomings of AI.
arXiv:2607.10916v1 Announce Type: new
Abstract: This study investigates how the coupling of two unit cells belonging to the same mechanical metamaterial into a dual unit cell configuration, can produce a new metamaterial with enhanced wave attenuation capabilities. For two metamaterials, two different unit cell coupling configurations are examined in 2D (side by side and chessboard), with particular emphasis on maintaining a plane crystallographic group of high symmetry, in order to simplify band structure calculations given the complexity of the geometry. It is shown that for specific configurations and choices of unit cell, multiple directional and/or omnidirectional band-gaps can appear, some of which can exhibit enhanced attenuation. The way in which these band-gaps emerge is described through applying the same procedure on 1D spring mass chains. Results support the idea that any band-gap metamaterial could have a much more efficient version which can be constructed purely from its own unit cells.
arXiv:2607.10941v1 Announce Type: new
Abstract: Monadic dependence is a proposed structural dividing line for fixed-parameter tractability of first-order model checking on hereditary graph classes. A graph class is \emph{monadically dependent} if the class of all graphs cannot be interpreted in its vertex-colored members using a fixed first-order formula. We prove two structural consequences of monadic dependence. First, every monadically dependent class has \emph{almost linear neighborhood complexity}: for every graph $G$ in the class and every set $A\subseteq V(G)$, the family $\{N_G(v)\cap A : v\in V(G)\}$ has size $|A|^{1+o(1)}$. Second, every $n$-vertex graph in a monadically dependent class has radius-1 merge-width $n^{o(1)}$. Here, merge-width is the decomposition parameter of Dreier and Toru\'nczyk based on construction sequences; its radius-$r$ version measures local reachability among parts through already resolved pairs. This settles the radius-1 case of the conjectured connection between monadic dependence and almost bounded merge-width and provides the first decomposition-based structural description of monadically dependent graph classes. Our proof is algorithmic: we give an $\mathcal{O}(n^5)$-time algorithm that, given an $n$-vertex graph $G$ such that $|\{N_G(v)\cap A : v\in V(G)\}|\le O(|A|^d)$ for every $A\subseteq V(G)$, computes a construction sequence witnessing radius-1 merge-width $\mathcal{O}(n^{1-1/d}\log n)$.
arXiv:2506.02161v3 Announce Type: replace
Abstract: The rapid advancements of Text-to-Image (T2I) models have ushered in a new phase of AI-generated content, marked by their growing ability to interpret and follow user instructions. However, existing T2I model evaluation benchmarks fall short in limited prompt diversity and complexity, as well as coarse evaluation metrics, making it difficult to evaluate the fine-grained alignment performance between textual instructions and generated images. In this paper, we present TIIF-Bench Text-to-Image Instruction Following Benchmark), aiming to systematically assess T2I models' ability in interpreting and following intricate textual instructions. TIIF-Bench comprises 5,000 prompts organized along multiple dimensions and categorized into three levels of difficulty and complexity. To rigorously evaluate robustness to prompt length, each prompt is provided in both short and long versions with identical core semantics. We further propose a novel Global Normalized Edit Distance (GNED) metric for text rendering and provide aspect-ratio-diverse reference images for each prompt to assess style control. In addition, we collect 100 high-quality designer-level prompts covering diverse scenarios for comprehensive evaluation. To enable scalable and fine-grained evaluation, we explore the best paradigm for leveraging the world knowledge encoded in large Vision-Language Models (VLMs) as automated binary evaluators. Through extensive ablations, we develop a fully reproducible evaluator that provides interpretable reasoning and reliable verification, enabling our benchmark to discern subtle variations in T2I model outputs. Through comprehensive benchmarking of mainstream T2I models on TIIF-Bench, we analyze the strengths and weaknesses of current T2I systems and reveal the limitations of existing evaluation benchmarks. Project Page: https://a113n-w3i.github.io/TIIF_Bench/.
arXiv:2607.10198v1 Announce Type: new
Abstract: Search APIs are the fundamental retrieval layer for many agents and are often their most frequently used tool. Traditional search APIs provide URLs, titles, and snippets that preview website contents. Because full-page retrieval is token-intensive, agent retrieval architectures increasingly use progressive disclosure: the agent first sees snippets and then chooses whether to fetch full pages. In such systems, search API performance is often evaluated primarily by answer accuracy. We argue that a commercial search API is better understood as a decision surface: the ranked snippets, URLs, and metadata that determine whether an agent answers immediately, searches again, or spends tokens opening pages. We test this claim with one frozen GPT-5.4 agent, two tools (search_web and fetch_page), and 100 questions from SEALQA-HARD, varying only the search provider (Brave, Tavily, Firecrawl). A Kimi-K2.6 oracle labels every content element visible to the agent (URL, title, snippet, and fetched page, when fetched), producing 6,869 valid per-URL judgments. We use an audited correct-answer label, semantic match, which preserves exact matches while accepting harmless formatting and naming variants. Under this measure, the providers remain close (25, 25, 26 / 100), but their evidence economies differ sharply: Brave offers gold-answer-rich snippets, Tavily concentrates gold-supporting URLs at rank 1, and Firecrawl is associated with broader exploration under this fixed agent policy. We also introduce a surface contradiction-to-gold URL ratio, which varies from 0.92 to 2.59. Provider choice is therefore a retrieval-budget and policy decision, not merely a recall decision.
arXiv:2607.09737v1 Announce Type: new
Abstract: Molecular docking predicts how a small molecule binds to a protein and is a key bottleneck in drug discovery. Classical scoring functions sum empirical pairwise contacts, blind to quantum-mechanical effects like orbital charge transfer that govern binding specificity. We introduce Q-Score, encoding GNN-predicted orbital donor-acceptor energies into a weighted graph and scoring binding by solving a maximum-weight vertex clique problem via Digitized-Counterdiabatic QAOA. Each interaction anchor maps to one qubit and compatibility constraints become edges. Across 11 protein targets, DC-QAOA recovers the exact optimum on 8 at 10 qubits. On 1000 AI-generated molecules, Q-Score is orthogonal to classical scoring with Spearman rho of 0.05, driven by orbital quality with rho of 0.90, and free of molecular-weight bias, enriching for strong orbital interactions at twice the random rate. DC-QAOA achieves a mean approximation ratio of 0.94 with 52 percent exact. Execution of 1000 circuits on IBM Eagle confirms 6-qubit solvability on NISQ hardware.
arXiv:2607.10201v1 Announce Type: new
Abstract: Equitable renewable-energy planning is a sequential decision problem, but the decision variables available to a public planner differ sharply between mature and emerging economies. In the former the government largely builds generation, while in the latter it steers private investment through incentives and quotas. We formulate socially-equitable renewable-energy budget allocation as a Markov Decision Process (MDP) and, using a single problem-agnostic solver interface, compare the same policies across the two settings: eight U.S. cities (a mature economy) and West Java, Indonesia (an emerging economy). The results show that across both settings, a receding-horizon value-iteration policy dominates. In the U.S., it reaches 66% renewable penetration while cutting the underserved low-income population by 96% versus a random baseline. In West Java it closes the low-access gap while crowding in the most private capital. More interestingly, a naive market-chasing heuristic, which is mildly sub-optimal in the U.S., could yield catastrophic outcomes in Indonesia, by underserving every low-access region, because chasing attractive markets and serving the underserved goals diverge once the planner acts through private developers.
arXiv:2506.07040v4 Announce Type: replace
Abstract: We study model-free methods for distributionally robust infinite-horizon average-reward Markov decision processes (MDPs). We present non-asymptotic convergence analyses of Q-learning and actor-critic algorithms for robust average-reward MDPs under contamination, total-variation distance, and Wasserstein uncertainty sets. A key ingredient of our analysis is showing that the optimal robust Bellman operator is a strict contraction with respect to a carefully designed semi-norm. This property enables a stochastic approximation update that learns the optimal robust $Q$-function with $\tilde{\mathcal{O}}(\epsilon^{-2})$ dependence on the target accuracy. We also establish robust TD convergence bounds whose constants are uniform over all stationary policies, yielding an efficient data-driven routine for robust critic estimation. Building on this, we introduce an actor-critic algorithm that learns an $\epsilon$-optimal robust policy with $\tilde{\mathcal{O}}(\epsilon^{-2})$ dependence on the target accuracy. We provide numerical simulations to illustrate the qualitative behavior of the proposed algorithms. Our results contribute to the theoretical foundations of robust planning under model misspecification and to model-free approaches for building robust long-run policies directly from simulation data.
arXiv:2607.09821v1 Announce Type: cross
Abstract: Medical imaging generates high-resolution images posing significant storage, transmission, and computational challenges. While low-rank matrix approximation (LoRMA) techniques offer efficient compression by exploiting structural redundancy, global approaches often fail to preserve local details critical for diagnosis. This paper focuses on clustering techniques that exploit non-local self-similarity to identify structurally similar regions in medical images. These clusters can be used for post-processing tasks such as adaptive image compression. We evaluate five clustering techniques: k-means, mini-batch k-means, agglomerative hierarchical clustering, balanced iterative reducing and clustering using hierarchies (BIRCH), and bisecting k-means across MRI, ultrasound, and chest X-ray modalities. All clustering techniques were optimised using random search, and cluster quality was assessed using the Silhouette score, the Davies-Bouldin (DB) index, and the Calinski-Harabasz (CH) index. Results demonstrate that standard k-means and bisecting k-means generally achieve strong cluster cohesion and separation across modalities. However, they tend to form a small number of clusters with high intra-cluster variability, limiting their effectiveness for post-processing tasks such as adaptive compression. Agglomerative clustering outperformed other techniques for MRI and ultrasound in terms of intra-cluster homogeneity, making it more suitable for preserving fine diagnostic details. For chest X-rays, mini-batch k-means achieved the best balance between clustering quality and intra-cluster compactness. BIRCH consistently underperformed across all modalities.
arXiv:2607.11272v1 Announce Type: cross
Abstract: Accurate dengue forecasting is crucial for public health planning, but remains challenging because incidence series are often short, noisy, non-stationary, nonlinear, and often affected by long-range temporal dependence. Fractional differencing in Autoregressive Fractionally Integrated Moving Average (ARFIMA) helps balance non-stationarity and persistence, but its linear structure limits its ability to capture nonlinear dynamics. Deep neural networks can model nonlinear patterns, but usually require large training samples and do not explicitly encode statistical long memory. Echo State Networks (ESNs), a widely used reservoir computing framework, are attractive in this setting because they retain nonlinear recurrent dynamics while training only a simple readout, making them suitable for data-scarce scenarios. However, standard ESNs lack long-term memory from a time-series perspective. This study proposes a long-memory reservoir computing framework that integrates dedicated long-memory and short-memory ESN reservoirs with a ridge-regression readout. We introduce two variants: Fractional ESN (fESN), which incorporates fractional-differencing dynamics into the reservoir to encode long-range dependence directly, and Wavelet ESN (wESN), which extracts stable low-frequency components through wavelet smoothing before modeling them with a memory-aware reservoir. We establish theoretical guarantees for closed-loop reservoir dynamics, showing that standard ESNs induce short-memory processes under mild conditions, whereas the proposed long-memory reservoirs generate polynomially decaying dependence consistent with statistical long memory. Across multiple dengue datasets and forecasting horizons, fESN and wESN outperform statistical and deep learning baselines. Combining conformal prediction with fESN and wESN provides distribution-free calibrated uncertainty intervals.
arXiv:2607.11065v1 Announce Type: new
Abstract: Detecting muscle fatigue via surface electromyography (sEMG) is essential for applications in sports, rehabilitation, and wearable health monitoring. Accurate and timely detection of fatigue is crucial for preventing injuries, optimizing physical performance, and ensuring user safety during prolonged activity. However, existing deep learning models are often unsuitable for this task due to their high computational cost and dependence on large-scale data. In this work, we propose an energy-efficient framework for muscle fatigue detection based on Spiking Neural Networks (SNNs), which exploit sparse, event-driven computation and temporal modeling. We further introduce a quantization-compatible training scheme (SDH) that combines multiple regularization terms to improve robustness under noisy conditions. Evaluated on two public sEMG datasets against a broad set of baselines and under seven noise conditions including physically motivated perturbations, our quantized SNNs match or exceed strong baselines while remaining more stable under diverse noise and reducing estimated energy consumption by up to 201.77x. These results demonstrate the framework's strong potential for real-time deployment in low-power wearable systems.
arXiv:2607.10208v1 Announce Type: new
Abstract: Accurate meteorological forecasting is essential for agricultural planning, irrigation management, and environmental decision support. This study conducts a comparative evaluation of recurrent and hybrid deep learning architectures for multivariate forecasting of reference evapotranspiration ($ET_0$), vapour pressure deficit (VPD), wind speed, and the sine and cosine components of wind direction. The analysis utilizes 134,376 hourly observations from Ioannina, Greece, spanning January 2011 to April 2026, sourced from ERA5 via the OpenMeteo Historical Weather API. Single and multi-layer GRU and LSTM networks are compared with hybrid 1D-CNN-GRU and 1D-CNN-LSTM models for two forecasting tasks: a 24-hour next-day forecast and a 168-hour week-ahead forecast. Performance is evaluated using normalized root mean squared error, the coefficient of determination, and a composite Weighted Quotient Score (WQS). The most effective purely recurrent models are a 64-unit LSTM for the 24-hour horizon, with a WQS of 0.816755, and a 1024-unit GRU for the 168-hour horizon, with a WQS of 0.779465. The hybrid CNN-GRU models achieved the highest overall scores of 0.827535 and 0.782863 for the 24-hour and 168-hour horizons, but with additionally more number of units respectively to LSTM models, while the CNN-LSTM models yield nearly identical results with substantially fewer parameters. Compared to the corresponding recurrent baselines, the hybrid models improve WQS by 1.22--1.63\% at 24 hours and by 0.44--0.45\% at 168 hours, indicating that convolutional feature extraction is more beneficial for short-term forecasting.
arXiv:2607.10915v1 Announce Type: new
Abstract: We introduce NAILS (Normative Alignment of Recommender Systems via Internal Label Shift), a simple and scalable method for aligning recommendation outputs with target distributions over item-level attributes, such as categories. Recommender systems optimized solely for user engagement often fail to satisfy broader normative objectives, including fairness, diversity, and editorial values. NAILS modifies the user-conditional item distribution to induce a specified marginal distribution over attributes while preserving the preferences learned by an existing recommender system and requiring no model retraining. We formulate this problem as a form of label shift applied internally within a hierarchical classification framework. By adopting a stakeholder-centric perspective, NAILS enables recommendation outputs to be aligned with global normative objectives. Empirically, we show that NAILS consistently improves attribute-level alignment with minimal impact on user engagement, providing a practical mechanism for value-driven recommendation.
arXiv:2607.09887v1 Announce Type: new
Abstract: Prophet inequalities have become a central tool for analyzing the performance of online algorithms. However, most existing results assume that input random variables are independent, which limits their applicability. Motivated by this gap, we study prophet inequalities under two correlation models induced by a latent state of the world variable $Z$. In the common-base model, the algorithm observes the sequence $Z+X_1,\dots,Z+X_n$. We analyze single-threshold algorithms with the constraint that they always accept the final item, guaranteeing a reward of at least $Z$. When $Z$ is chosen adversarially, we characterize the optimal deterministic algorithm of this form, achieving a competitive ratio of $0.381$. We then show that randomizing improves the guarantee to $0.4$. By a minimax argument, the same ratio is achievable when $Z$ is random.
We depart from standard techniques by establishing a stronger lower bound of $0.41$ and an upper bound of $0.475$, ruling out the possibility that this class of algorithms attains the $1/2$ ratio known for independent inputs. The core technical contribution is a new analytical framework that captures the reward dynamics of single-threshold algorithms. We introduce a differential equation characterizing the expected reward of a threshold in the worst-case instance, parameterized by the distribution of the maximum. This equation admits a closed-form and unifies known single-threshold prophet inequalities, yielding a simple threshold-optimality condition applicable to the common-base model.
Finally, we study the common-scale model, where inputs take the form $Z\cdot X_1,\dots,Z\cdot X_n$. We show that this minimal multiplicative correlation yields strong impossibility results: no algorithm can achieve a competitive ratio exceeding $1/n$.